Skills
skills/gracefullight/stock-checker/deepsec/Gen Agent Trust Hub

deepsec

Fail

Audited by Gen Agent Trust Hub on Jun 14, 2026

Risk Level: HIGHDATA_EXFILTRATION
Full Analysis
  • [DATA_EXFILTRATION]: Path traversal sequence detected attempting to access internal system directories.
  • Evidence: ../../../.agents/workflows/deepsec.md in SKILL.md.
  • The skill uses multiple parent directory references (../) to escape its own directory structure and target the .agents/workflows/ directory. This directory typically contains internal configuration and sensitive workflow definitions for the agent software, and unauthorized access to these files can lead to data exposure or manipulation of the agent's core behavior.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Jun 14, 2026, 11:38 PM
Security Audit — agent-trust-hub — deepsec